[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[Xmca-l] Re: units of analysis?

Seems to me the problem in many research projects is that the question is not formulated in an appropriate way. LSV was exploring a method of analysis that seeks to understand the relationship among components in a complex totality. Not the causal influence of one factor on another, which is often how students frame their research interest. And this means that the unit of analysis has to represent, exemplify, this relationship.


On Oct 12, 2014, at 1:31 PM, Helena Worthen <helenaworthen@gmail.com> wrote:

> As someone who uses the concept of "unit of analysis" in a very down-to-earth, quick and dirty, applied way to shape collective responses to a crisis in a labor and employment relationships (like, when a rule changes creates difficulties for workers), I would agree with Andy:
>> The other thing is that discovering a unit of analysis is an *insight*. It
>> is not something that can be achieved by following a template, it is the
>> breakthrough in your research into some problem, the leap. It usually comes
>> *after* you've collected all the data for your research using some other
>> unit of analysis.
> First comes the story, the details, the experiences. The question lying behind the telling of the stories is, "What are we going to do?" The unit of analysis gets defined by the purpose we are trying to accomplish. Are we trying to get the employer to back off temporarily? Are we trying get the rule changed? Example:  In a big hospital system in Chicago, clerical workers were no longer allowed to leave an "I'm going to be late to work today" or "I have to stay home with my sick kid today and will miss work" message on the answering machines of their supervisors. We're talking about a workforce with hundreds of employees, most of them middle aged minority women -- with grandchildren and extended families to be responsible for.  Not being allowed to leave a message on a machine, but being required to actually speak to a supervisor in person who would then keep a record of the call, was a problem because supervisors were often away from their desks and the whole phone system was unreliable. Also, a lot of workers didn't have cell phones at the time this was happening (2004) and pay phones are few and far between, so if someone it out buying more asthma inhalers for a grandkid, making a phone call is not easy.
> So, exactly what is the purpose that we're trying to accomplish, here?  To repeal the rule? To fix the phone system?  To educate members of the union and other others about how to respond collectively to something that only affects some of them? To make a profound change in society so that middle-aged women are not the primary caretakers of an extended family?  Pick one. Once you've picked one (hopefully, one that you can carry out) you can define the unit of analysis and then reviewing the whole Engestrom triangle and evaluating your strategy becomes, as Andy says,  a matter of solving puzzles. 
>> From the employer point of view, asking workers to actually speak to a live supervisor makes a certain sense. That's why we talk about activity system(s), not just one activity system. But they are often in conflict with each other, which adds to the drama.
> Is the data in your study being gathered with some purpose in mind? Is the purpose the purpose of the children, the purpose of the class, or the purpose of the PhdD program?  To me, what would be most interesting would be a comparison between the unit of analysis (purposes of children) and unit of analysis (purpose of classroom). I'll bet they're not identical.
> Helena 
> Helena Worthen
> helenaworthen@gmail.com
> On Oct 12, 2014, at 10:20 AM, Katerina Plakitsi wrote:
>> This problem of the ' unit of analysis' is my concern too. I supervise
>> three PHD students on Science Education in a CHAT context. Two of them on
>> early childhood science education and one on primary science. They have
>> collected log files, children discourses consisted of
>> scientific justifications, accepted rules, and forms of division of labor.
>> They have collected children narratives, and drawings. When they decided to
>> analyze their data they follow different paths into CHAT context mainly
>> modeling them using Engestrom's triangle. They still doubt about the " unit
>> of analysis".
>> Στις Κυριακή, 12 Οκτωβρίου 2014, ο χρήστης Andy Blunden <ablunden@mira.net>
>> έγραψε:
>>> Katie, picking up on your concern about units of analysis, it was one of
>>> the points I mentioned in my "report" from ISCAR, that this concept was
>>> almost lost to us. A phrase I heard a lot, and which was new for me, was
>>> "unit to be analysed." If anyone knows the origin of this expression, I'd
>>> be interested in hearing. It seemed to me that what it referred to was a
>>> "closed system" for analysis, that is, abandoning CHAT methodology whilst
>>> keeping the word. If I am mistaken about this, please let me know.
>>> The other thing is that discovering a unit of analysis is an *insight*. It
>>> is not something that can be achieved by following a template, it is the
>>> breakthrough in your research into some problem, the leap. It usually comes
>>> *after* you've collected all the data for your research using some other
>>> unit of analysis. In Kuhn's terms, discovery of the unit is the new
>>> paradigm, after which it is just a matter of solving puzzles. So for
>>> graduate students to use the concept of unit in their research, often
>>> depends on the wisdom of teh direction they get from their supervisor. I
>>> don't know how many PhD students I've met who have got to this point in
>>> their thesis and discover that the data they have is not the data they now
>>> know they need.
>>> Andy
>>> ------------------------------------------------------------------------
>>> *Andy Blunden*
>>> http://home.pacific.net.au/~andy/
>>> Katherine Wester Neal wrote:
>>>> I like Holli's plan to commit some time to reading the two articles. But,
>>>> as usual, I don't know that I'll have much to contribute in posts. I
>>>> usually get deep in thinking about the posts and don't follow that through
>>>> to write something. The writing is much harder, and I am usually just
>>>> trying to keep up with reading!
>>>> For me, the thread has been fascinating, probably because I'm interested
>>>> in different units of analysis, what they might be used for, and how they
>>>> fit together with theory and conducting research. What are people doing
>>>> with units of analysis and why? Or why aren't units of analysis being used?
>>>> If anyone wants to write more in that direction, I'd be very interested to
>>>> read, and I'll try to respond, although the questions might be as basic as
>>>> these.
>>>> Lastly, Andy has basically been articulating my thoughts (in a much more
>>>> eloquent way than I would) about action as a unit of analysis. In Mike's
>>>> example about driving and thinking and writing, I'd add that the action is
>>>> mediated. Together with sociocultural and historical factors that
>>>> influenced those actions (and which, as has been said here before, are
>>>> often difficult to get a look at), the actions create a picture of much
>>>> more than just Mike's behavior.
>>>> Katie
>>>> Katie Wester-Neal
>>>> University of Georgia
>> -- 
>> ............................................................
>> Katerina Plakitsi
>> Associate Professor of Science Education
>> School of Education
>> University of Ioannina
>> University Campus Dourouti 45110
>> Ioannina
>> Greece
>> tel. +302651005771
>> fax. +302651005842
>> mobile.phone +306972898463
>> http://users.uoi.gr/kplakits
>> http://erasmus-ip.uoi.gr
>> http://www.lib.uoi.gr/serp